Association for Public Policy Analysis & Management

Panel: Implications of Health Insurance Coverage Reporting Accuracy

(Health Policy)

Survey estimates of health insurance coverage are important to understanding access to health care, especially as new policies are introduced. Prior research indicates that survey estimates of insured/uninsured status were reasonably accurate, but estimates of particular types of coverage (Medicaid, non-group insurance, etc.) have measurement error. Deriving accurate estimates of coverage has become even more complicated with the introduction of new plans and pathways to gaining health insurance. Specifically, private coverage with tax credits and public plans that require a monthly premium that are both available along with traditional Medicaid through the same “no wrong door” pathway, namely the Marketplace or Healthcare.gov. To gain a more comprehensive understanding of this measurement error, a validation study called CHIME (Comparing Health Insurance Measurement Error) was launched. Using administrative records from a private Minnesota health plan, individuals known to be enrolled in a range of different coverage types (including employer-sponsored insurance, non-group coverage, qualified health plans from the Marketplace, and public coverage) were selected and randomly assigned to one of two survey treatments: the Current Population Survey (CPS) and the American Community Survey (ACS). The purpose of this panel is to provide an overview of CHIME validation study’s motivation and methodology, provide key study results, and discuss the implications of the results for using survey estimates of health insurance coverage for evaluating the impact of federal and state health reform efforts. Specifically, we examine the magnitude, direction and patterns of under-reporting health insurance status and coverage type in both Census Bureau surveys commonly used to inform and evaluate health policy. We also explore the offsetting influence of misreporting coverage across individuals with different insurance types (e.g., are reports of Medicaid among individuals known to have non-group coverage offset by reports of non-group coverage among individuals known to have Medicaid). Finally, we project the impact of coverage type misreporting for estimates of the prevalence of insurance coverage derived from state-specific and national benchmark data. The first presentation provides an overview of what is known about accurate reporting of insurance coverage, a description of the CHIME methodology and insurance type categorization rules, and an overview of the research questions and analysis plan. The second and third presentations show how accurately people report insurance status (insured/uninsured), and how accurately people are classified as into the correct insurance types, respectively, for the CPS and ACS survey treatments. The CPS and ACS presentations will also show the impact of misreporting on population estimates of coverage. Finally, the discussant will compare and contrast key findings from the presentation of CPS and ACS results. Most importantly, the discussion will center on the implications of CHIME for those working with state and federal surveys and data. Specifically, what are the implications of CHIME results for survey design and for what is known about estimates of the distribution of health insurance coverage derived from federal surveys (e.g., ACS, CPS and the National Health Interview Survey (NHIS)) that are routinely used to monitor and evaluate the impact of health policies.